Click here to load reader
Upload
truongdiep
View
212
Download
0
Embed Size (px)
Citation preview
Abstract—This paper presents localization in wireless sensor
networks (WSNs) in presence of obstacles using a mobile
anchor. The proposed scheme uses a mobile anchor, which finds
its location via global positioning system (GPS). The mobile
anchor broadcasts its coordinate while it navigates through
sensing region. The static sensor nodes receive beacon messages
when the mobile anchor and sensor nodes have overlapping
signals. The sensor nodes process the beacon messages to find
their own coordinates using location of the anchor. The anchor
moves in a snake-like method, which consumes less energy
compare to the other approaches such as random methods. In
this method the need for specific ranging hardware is
eliminated on the sensor nodes. This elimination voids the
communication requirement between sensor nodes. Further,
the presence of obstacles in the environment has almost no
effect on the output of the system causes the obstacle can be
detected by the mobile anchor. In their project, Castalia
simulator is used for the simulation and validation purpose. The
output of the simulation results of localization problem confirm
less localization error and less energy consumption over other
range free solution using mobile anchor considering an
extensive variety of network conditions.
Index Terms—Localization, mobile anchor nodes, wireless
sensor networks, obstacle environment.
I. INTRODUCTION
Localization wireless sensor network has been point of
interest in recent years in many applications. With the
ongoing revolution in the hardware and wireless technologies,
wireless sensor network is becoming the most preferred
option for many applications, e.g., environment monitoring,
health care, wildlife habitat monitoring, and tracking of
objects [1]. Sensor nodes are small devices, which equipped
with functions for environmental sensing and wireless ability.
A set of sensor nodes can create a wireless sensor network.
Localization problem of these sensors is an interested area in
WSNs due to its crucial information. In localization problem
the researcher needs to find accurately the location of
individual sensor.
In range-free scheme for localization, which is an
interested scheme for many applications of WSNs, the
accuracy is lower compare to the rage-based scheme but in
comparison the cost of the hardware is lower than the one in
range-based [2].
This study proposes a range-free scheme for locating static
Manuscript received January 7, 2013; revised April 16, 2013.
Sayyed Majid Mazinani is with the Iranian Construction Engineering
Organization Province of Khorasan Razavi and Department of Electrical
Engineering, Imam Reza University Mashhad-Iran (e-mail:
Fatemeh Farina is with the Department of Computer Engineering of Azad
University Mashhad-Iran (e-mail: [email protected]).
sensors by using a mobile anchor in presence of obstacles.
The static sensors are in a two-dimensional coordinate
system, which are distributed in sensing region. A mobile
anchor, which has a snake-like manner, can navigate through
entire regain. The strength point of this research is that the
anchor has ability to face the obstacles and circulate them
while cover the entire surface visiting only once every place
passing any place more than once. The purpose of the
research mostly focuses the less error of detecting the
location and lower energy consumption of mobile anchor.
The remaining of the paper proceeds as follows. Section II
presents the literature review of localization scheme. The
proposed methodology of localization is presented in Section
III. Next, Section IV presents the simulation and discusses
the results of simulation. Finally, Section V concludes and
presents the limitations of this research.
II. LITERATURE REVIEW
Localization in wireless sensor network has been
interested in different kind of situations. Accuracy, cost, and
scalability are three main factors to be considered when
designing localization method with high energy efficiency.
Table 1 shows the brief review of the literature in localization
problem of wireless sensor network having different
conditions.
[3] develops a Range-Free localization technique with
mobile anchors equipped with GPS. Their method offers
coarse positioning accuracy with low computational expense.
Conversely the energy consumption of their scheme as the
mobile anchors moves randomly.
Kim et al. developed a new range-based localization
scheme of mobile beacon-assisted localization (MBAL) [4].
Their technique has simple computational algorithm for
movement path selection. Their scheme also uses a new
range-checking technique for position-ambiguity problem.
Although this scheme offers more promising results over
random movement method, it fails to have the lowest energy
consumption then compare to other schemes.
Li et al. used graph theory for the path planning of mobile
anchors [5]. They considered the wireless network sensor as a
connected undirected graph; then they redefined the
path-planning problem into traversing graph and Spanning
Tree with Breadth-First (BRF) and Backtracking Greedy
(BTG) algorithms for the latter.
Karim et al. developed a Range-free localizing scheme by
using Mobile Anchor (RELMA) for large scale WSNs,
offering higher accuracy and energy efficiency. However the
movement of the mobile anchors is random, leading to more
energy consumption [6].
Chen et al. proposed a localization technique based on ring
Localization in Wireless Sensor Network Using a Mobile
Anchor in Obstacle Environment
Sayyed Majid Mazinani and Fatemeh Farnia
438
International Journal of Computer and Communication Engineering, Vol. 2, No. 4, July 2013
DOI: 10.7763/IJCCE.2013.V2.222
overlapping with reference and blind nodes [7]. The anchor
sensor node can move snake-like for localization static sensor
nodes. They implement Received Signal Strength
Indicator(RSSI) values to localize the rings including
reference and blind nodes. By overlapping the rings with
each other and by utilizing the location of reference nodes,
they implemented an algorithm to localize the blind nodes.
Ou presented a location scheme in WSN applying mobile
anchors with four directional antennas (DIR) [3]. However
the movement of mobile anchor in the simulation is straight
line, which may path a certain place more than one time,
which cause higher energy consumption.
One of the advantages of proposed simulation over other
simulation is that it needs minimum two beacon messages for
locating static sensor nodes, while in many proposed models
the minimum number of messages is three. Lower
transmitting information in simulation means less cost which
is the priority in localization of WSNs.
III. PROPOSED APPROACH
In this section we explain the system environment of the
localization problem in wireless sensor networks. Following
the anchor mobile characteristics and the method of finding
the location of sensor nodes are presented. The purpose of
this research is to present a robust method for estimating the
location of sensors using a mobile anchor with low energy
consumption in obstacle environment.
Localization parameters of our method are beacon
messages.
A. System Environment
In the simulation there are two types of sensors including
static sensors nodes and a mobile anchor node. Static sensors
nodes are randomly distributed in a two-dimensional
coordinate system. The locations (states) of static sensors are
unknown since they do not use GPS receiver. The mobile
anchor node, which is equipped with an Omni directional
antenna, employs a GPS receiver to discover their locations
when they navigate over the sensing field. The anchor is able
to receive a message from sensor nodes while traveling
through region.
B. Beacon Messages
Localization parameters of our method are beacon
messages. In the proposed model the anchor node broadcasts
messages called beacon messaged while navigating in the
sensing area. The beacon message contains time of
messaging which is a sequential number of message and its
coordinate in the system. The sequence number and its
location the anchor changes while it moves.
TABLE I: COMPARISON OF RELATED WORKS IN LOCALIZATION SCHEME FOR WSNS USING MOBILE ANCHOR
Ssu MBAL Path
planning RELMA
Ring
overlapping
Directional
antenna
Proposed
method
Centralized /Distributed distributed
central
Scalability x x x x x x x
Range based/free Range free Range based Range free Range free Range free Range free Range free
mobile anchor movement Random Path decision Graph Random Snake-like Straigh line Snake-like
environment obstacle x
Number of mobile anchors 1 1 1 n 1 n 1
Neighbor-information-based x x
Directional antenna x x
Energy consumption high medium low high medium high low
minimum number of
messeages for localization 4 3 3 3 4 6 2
C. Movement
It is essential to reduce the path of mobile anchor and to
have full coverage in order to save the hardware cost of
sensor nodes. Therefore we define the movement of the
anchor sensor node in a snake-like method to deploy static
sensor node. As we see in Fig. 1, it shows the
two-dimensional coordination system of WSNs existing one
obstacle. Anchor begin from the corner of the sensing field
(starting point is (0,0)) to navigate through entire area
moving from left to right, moving down, then moving from
right to left, again moving down and so on. An obstacle can
be circulated by the snake-like method, since the mobile
anchor change its direction while it faces the obstacle. When
the anchor changes its direction it stores the information of
circulating point for future movements to have correct
direction. The mobile anchor transmits beacons periodically
while they are travelling to set the coordinate for each static
sensor nodes in turn. Accordingly, sensor nodes can cover
every area in the sensing field while they don't travel any area
more than once. This reduces the energy consumption of the
mobile anchor while passing the entire region.
D. Location Estimation
While mobile anchor broadcasting the beacons
sequentially, the sensor nodes can receive the beacon
messages when the anchor is in transition neighbor. Each
sensor node is able to calculate its location using two beacon
messages. We assume that the coordinates of two different
locations of mobile anchor is (x1, y1) and (x2, y2). Each sensor
node calculates its coordinate according to (1) and (2).
(x-x1)2+(y-y1)
2=r2 (1)
(x-x2)2 + (y-y2)
2=r2 (2)
Fig. 2 shows the location estimation of a sensor node. As
we see it reflects two estimated locations for each sensor
node. For finding the best solution the anchor uses the degree
of antenna.
439
International Journal of Computer and Communication Engineering, Vol. 2, No. 4, July 2013
Fig. 1. Snake-like pattern of the mobile anchor facing an obstacle.
IV. SIMULATION AND ANALYSIS
In this section the simulation environment, the
performance of the proposed simulated model, and the results
including the comparisons of different factors are analysed
and presented.
Simulation
For simulating the proposed model, Castalia based on
Omnet ++ is employed. In this simulation, the sensor region
is considered as 10,000 m2 while there are 100 sensor nodes,
which randomly distributed, and one mobile anchor. As we
see in Fig.1, we define the initial position of the mobile
anchor as top-left of the sensor filed. Using Snake-like
pattern, it navigates through the entire region.
In our simulations, the average location error is the average
distance between estimated location (Xei, Yei) and the actual
location (Xi, Yi) of all sensor nodes. The average location
error is defined in (3).
We consider some parameters for showing the efficiency
of the presented work. The simulation parameters are shown
in Table II.
Anchor locations
Actual location of sensor
node
Estimated locations
Fig. 2. Location estimations of a sensor node.
In our simulation we also consider the existence of
obstacles as we also have this situation in real-word.
Therefore we placed several obstacles in the sensing region to
can have more realistic results.
TABLE II: SIMULATION PARAMETERS
Parameters Values
The impact of Beacon distance
on average localization error 1, 2, 3, 4, 5 m
The impact of Beacon distance
on energy consumption 1, 2, 3, 4, 5 m
V. RESULTS
In this section we consider some parameters to compare
the results of this simulation to the other similar presented
localization schemes. The most important factors to compare
are average localization error and energy consumption of the
anchor mobile.
A. The Impact of Beacon Distance on Localization Error
The impact of beacon intervals on the average localization
error is a method to measure the efficiency of the presented
localization scheme. To estimate the impact of beacon
intervals, we implement simulations with different beacon
intervals: 1m, 2m, 3m, 4m and 5m. We also compare our
method with DIR method, which is more similar to our work.
The simulation shows that the average localization error is
more or less similar to the DIR method. However, we also
compare the energy consumption of the anchor mobile of the
two different methods.
Fig. 3. Average localization error vs beacon distance.
B. The Impact of Beacon Distance on Energy Consumption
We also consider the energy consumption of the anchor
mobile. For comparing the results with DIR method, we
consider the impact of the bacon distance on the energy
consumption. We consider 1m, 2m, 3m, 4m and 5m as the
beacon messages intervals. Fig. 4. shows that the energy
consumption is lower in the presented method compare to
DIR method.
Fig. 4. Energy consumption vs beacon distance.
VI. CONCLUSION
Localization problem is one of the important problems in
many applications of Wireless Sensor Networks (WSNs). In
addition, range free problems are also more applicable
Ob
stac
le
440
International Journal of Computer and Communication Engineering, Vol. 2, No. 4, July 2013
2 2( ) ( )
number of sensor nodes
ei i ei ix x y y (3)
compare to range-based problems. The Simulation results
show that the average localization of static nodes has less
error compare to the other similar work. In our simulation we
also have one anchor mobile for localization sensor nodes
with minimum number of beacon transmissions, which cause
lower cost. Energy consumption of the anchor also plays
important role. We also consider this parameter to compare
other similar works. The simulation results shows that it
consume less energy using snake-like pattern in obstacle
environment.
REFERENCES
[1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, “A
survey on sensor networks,” IEEE Communications Magazine, vol. 40,
no. 8, pp. 102-114, August 2002.
[2] K. F. Ssu, C. H. Ou, and H. Jiau, “Localization with mobile anchor
points in wireless sensor networks,” IEEE Transactions on Vehicular
Technology, vol. 54, pp. 1186–1197, May 2005.
[3] C. H. Ou, “A localization scheme for wireless sensor networks using
mobile anchors with directional antennas,” Sensors Journal, IEEE, vol.
11, no. 7, pp. 1607-1616, 2011.
[4] K. W. Kim and W. J. Lee, “MBAL: A mobile beacon-assisted
localization scheme for wireless sensor networks,” in Proc. 16th
International Conference on Computer Communications and Networks,
2007. ICCCN 2007, IEEE, 2007.
[5] H. J. Li, J. W. Wang, X. Li, and H. X. Ma, “Real-Time path planning of
mobile anchor node in localization for wireless sensor networks,” in
Proc. 2008. ICIA 2008. International Conference on Information and
Automation, IEEE, pp. 384-389, 2008.
[6] L. Karim, N. Nasser, and T. E. Salti, “RELMA: A range free
localization approach using mobile anchor node for wireless sensor
networks,” in Proc. Global Telecommunications Conference
(GLOBECOM 2010), IEEE, 2010.
[7] Y. S. Chen et al., “An efficient localization scheme with ring
overlapping by utilizing mobile anchors in wireless sensor networks,”
in Proc. 2010 4th International Conference on Multimedia and
Ubiquitous Engineering (MUE), IEEE, 2010.
Sayyed Majid Mazinani was born in Mashhad, Iran on
28 January 1971. He received his Bachelor degree in
Electronics from Ferdowsi University, Mashhad, Iran in
1994 and his Master degree in Remote Sensing and
Image Processing from Tarbiat Modarres University,
Tehran, Iran in 1997. He worked in IRIB from 1999 to
2004. He also received his phD in Wireless Sensor
Networks from Ferdowsi University, Mashhad, Iran in
2009. He is currently assistant professor at the faculty of Engineering in
Imam Reza University, Mashhad, Iran. He was the head of Department of
Electrical and Computer Engineering from 2009 to 2012. His research
interests include Computer Networks, Wireless Sensor Networks and Smart
Grids.
Fatemeh Farnia was born in Mashhad, Iran in April
1978. She received her bachelor degree in computer
engineering from Azad university of Mashhad in 2010
where she is currently working toward M.S. degree in
department of computer engineering. Her current
research interests include wireless sensor networks,
localization and mobile system.
441
International Journal of Computer and Communication Engineering, Vol. 2, No. 4, July 2013